roberta_sst5_padding80model

This model is a fine-tuned version of roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 3.5247
• Accuracy: 0.5624

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.2916	1.0	534	1.3748	0.4484
1.0328	2.0	1068	1.0661	0.5317
0.8673	3.0	1602	1.1531	0.5195
0.7557	4.0	2136	1.1215	0.5692
0.6258	5.0	2670	1.2796	0.5367
0.528	6.0	3204	1.4486	0.5412
0.413	7.0	3738	1.5099	0.5511
0.3434	8.0	4272	1.6205	0.5552
0.2961	9.0	4806	1.7024	0.5394
0.2508	10.0	5340	2.1916	0.5398
0.2382	11.0	5874	2.5628	0.5357
0.1928	12.0	6408	2.6311	0.5534
0.1934	13.0	6942	2.9223	0.5489
0.157	14.0	7476	3.1200	0.5493
0.1235	15.0	8010	3.1288	0.5498
0.0967	16.0	8544	3.2570	0.5620
0.0911	17.0	9078	3.3236	0.5606
0.0698	18.0	9612	3.4772	0.5575
0.0595	19.0	10146	3.4699	0.5656
0.0579	20.0	10680	3.5247	0.5624

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.1
• Datasets 2.12.0
• Tokenizers 0.13.3